Journal of Physics Conference Series
OPEN ACCESS
Toward atom probe tomography of microelectronicdevicesTo cite this article D J Larson et al 2011 J Phys Conf Ser 326 012030
View the article online for updates and enhancements
You may also likeComparison of atom probe tomographyand transmission electron microscopyanalysis of oxide dispersion strengthenedsteelsA J London S Lozano-Perez S Santra etal
-
An atom probe tomography and inventorycalculation examination of second phaseprecipitates in neutron irradiated singlecrystal tungstenPhilip D Edmondson Baptiste Gault andMark R Gilbert
-
Behaviour of Nanostructured Si-BasedAnode Materials Studied By Atom ProbeTomography and Hard X-RayPhotoelectron SpectrocopyFrank Uwe Renner Konda GokuldossPradeep and Yueming Zheng
-
Recent citationsAtom Probe Tomography Characterizationof Dopant Distributions in Si FinFETChallenges and SolutionsRong Hu et al
-
Thomas F Kelly-
Nanoskalige chemische Bildgebung vonZeolithen durch AtomsondentomographieJoel E Schmidt et al
-
This content was downloaded from IP address 153177198113 on 02122021 at 1808
Toward atom probe tomography of microelectronic devices
D J Larson1
D Lawrence1 W Lefebvre
2 D Olson
1 T J Prosa
1 D A Reinhard
1
R M Ulfig1 P H Clifton
1 J H Bunton
1 D Lenz
1 J D Olson
1 L Renaud
3
I Martin3 and T F Kelly
1
1 Cameca Instruments Inc 5500 Nobel Drive Madison WI 53711 USA
2 Universiteacute de Rouen Saint Etienne du Rouvray 76801 FRANCE
3 Cameca SAS 29 Quai des Greacutesillons Gennevilliers 92622 FRANCE
Summary Atom probe tomography and scanning transmission electron microscopy has been
used to analyze a commercial microelectronics device prepared by depackaging and focused
ion beam milling Chemical and morphological data are presented from the source drain and
channel regions and part of the gate oxide region of an Intel
i5-650 p-FET device
demonstrating feasibility in using these techniques to investigate commercial chips
1 Introduction
Shrinking feature sizes continue to provide one of the main challenges for physical metrology
methods From the 2009 ITRS [1] ldquoTo achieve desired device scaling metrology tools must be
capable of measurement of properties on atomic distancesrdquo Atom probe tomography nominally has
this capability and has been used in materials characterization and materials science for more than 40
years [2] Although using a laser to assist field evaporation of materials in the atom probe was
developed ~30 years ago [3] only recently has APT begun to see widespread usage in the areas of
semiconductors [4-7] and ceramics [8-12] Although APT has also recently been applied to transistor
and FINFET-type structures [13-15] these structures usually have been stopped at some point in the
semiconductor build process in order to accommodate the APT analysis As it is a complex multi-step
task to analyze a fully processed microelectronic device using APT we will take a step back and
evaluate the three steps that are necessary to result in a successful APT analysis of any structure
Number one is specimen preparation For APT the use of focused ion beam (FIB) instruments
has revolutionized specimen preparation (for a recent review see [16]) For the case of microelectronic
device analysis FIB preparation to isolate lt50nm heterogeneous devices in arbitrary XYZ
orientations is critical This occurs only after suitable deprocessing (may consist of depackaging dry
and wet electrochemistry for selective material etching etc) has been successfully achieved Since a
protective layer is needed if near-surface features are to be analyzed proper ldquocap matchingrdquo becomes
important The important requirements are relatively low temperature deposition (to avoid inducing
undesired diffusion) matching of evaporation field with target material (to minimize reconstruction
artifacts) and good adhesion
Number two is specimen yield during data collection Even though a specimen may be prepared
which contains the region of interest specimen yield must be adequate (for the individual userrsquos
purpose or specific laboratory environment) Clearly the details of specimen preparation are
intricately related to yield but data collection parameters may also have a significant effect on yield
and there is often a trade-off between yield and data quality (primarily mass spectral and background
noise quality)
Davidlarsonametekcom
17th International Conference on Microscopy of Semiconducting Materials 2011 IOP PublishingJournal of Physics Conference Series 326 (2011) 012030 doi1010881742-65963261012030
Published under licence by IOP Publishing Ltd 1
Number three is
APT data
reconstruction which
is not discussed
significantly in this
document Data may
be collected with an
adequate yield but if
the resulting data
reconstruction is either
not accurate or not
precise enough then
the analysis may fail to
meet requirements
The main problem
with APT data
reconstruction from
the perspective of devices is that field evaporation of heterogeneous struc
evaporated surfaces that are far from hemispherical
fundamental assumption employed in the majority
This work presents the status
commercial microelectronics device (
we acknowledge that satisfactory yield has not
reconstruction of transistors are non
2 Experimental
Specimen preparation was carried out in an FEI Novalab
with an Omniprobe AP200 in-situ
[110] orientation of the Si substrate with a JEM ARM 200F operating at 200kV equipped with a
Schottky field emission gun APT data collection was performed on a LEAP
Cameca Instruments Inc The atom probe was operated in a 200 kHz pulsed laser mode with an
energy of 100 pJ into an estimated spot size (four sigma) at the specimen of ~3
temperature was 50K and the ion detection rate was 020 (1 ion detected in every 500 laser pulses)
The specimen examined in this work is
(Intel i5-650) which was purchased at a
retail outlet Following depackagi
(approximately) metal-1 [18] focuse
ion-beam preparation [19] was used to
create specimens Figure 1 shows
sequence of images throughout the
process from the initial surface of the
wafer after depackaging (figure 1a)
through to the final state of focused
beam annular sharpening [20] of the tip
(figure 1h)
3 Results and Discussion
The device structure analyzed with APT
is shown in figure 2 which contains a bright
figure 2a a multilayered gate oxide structure is
angled regions between the SiGeB sourcedrain regions and the channel become much more obvious
The APT analysis obtained from a device in a region near the one
Figure 1
APT analysis a) initial su
coupon d)
beam annular sharpening
is that field evaporation of heterogeneous structures often leads to
are far from hemispherical [17] and a hemispherical end form
assumption employed in the majority of current reconstruction algorithms
This work presents the status of efforts to prepare analyze and reconstruct data from
commercial microelectronics device (32 nm node Intel
i5-650 nFET) This is a work in progress and
we acknowledge that satisfactory yield has not yet been obtained Also issues relating to APT data
of transistors are non-trivial and will not be discussed further herein
Specimen preparation was carried out in an FEI Novalab dual-beam focused ion beam instrument
micromanipulator STEM observations were performed along the
e Si substrate with a JEM ARM 200F operating at 200kV equipped with a
APT data collection was performed on a LEAP 4000XH
The atom probe was operated in a 200 kHz pulsed laser mode with an
of 100 pJ into an estimated spot size (four sigma) at the specimen of ~3 microm The specimen
temperature was 50K and the ion detection rate was 020 (1 ion detected in every 500 laser pulses)
The specimen examined in this work is a device from a commercial 32-nm technology
was purchased at a
retail outlet Following depackaging to
focused-
was used to
1 shows a
sequence of images throughout the
the initial surface of the
(figure 1a)
focused-ion-
of the tip
with APT
ntains a bright-field high-angle annular-dark-field pair of images
gate oxide structure is clearly visible while in figure 2b the
sourcedrain regions and the channel become much more obvious
The APT analysis obtained from a device in a region near the one shown in figure 2 is
Focused ion beam preparation of a microelectronic device for
APT analysis a) initial surface b) FIB-milled trenches c) the extracted
coupon d) pre-tip cut from the coupon and e-h) the stages of focused
beam annular sharpening
Figure 2 (a) Bright-field and (b) high-angle annular
field images of the Intel i5-650 device
tures often leads to
hemispherical end form is a
and reconstruct data from a
650 nFET) This is a work in progress and
issues relating to APT data
focused ion beam instrument
STEM observations were performed along the
e Si substrate with a JEM ARM 200F operating at 200kV equipped with a
4000XHR from
The atom probe was operated in a 200 kHz pulsed laser mode with an
m The specimen
temperature was 50K and the ion detection rate was 020 (1 ion detected in every 500 laser pulses)
nm technology chip
field pair of images In
visible while in figure 2b the undercut and
sourcedrain regions and the channel become much more obvious
in figure 2 is presented
Focused ion beam preparation of a microelectronic device for
milled trenches c) the extracted
stages of focused-ion-
angle annular-dark-
17th International Conference on Microscopy of Semiconducting Materials 2011 IOP PublishingJournal of Physics Conference Series 326 (2011) 012030 doi1010881742-65963261012030
2
and gt500 (FWTM) which demonstrates
very complicated heterogeneous structures The Hf
3 in a region estimated to have a composition of approximately 80at (H
that the apparent HfO molecules detected in the channel region are spectral noise due to the large
number of mass ranges required for the HfO isot
the left and right edges of the image shown in figure 3 is ~25at
shown in figure 3 (arrowed) delineate
shape which resembles the undercut shape of the channel region near the gate oxide shown in figure
2b
This shape correlation may be used to create a compos
figure 5 in which the APT image was scaled to match the STEM image
of carbon atoms clustering together
concentration of carbon in the APT data is
~010at
Although these data are not from
exactly the same volume of material
(something which has been done very few
times to date in the literature [21]
exercise is still useful It provides us with
information on the accuracy of the APT da
reconstruction as well as perhaps a
of the future of correlative microscopy
Indeed TEM and APT are complementary
techniques that will likely develop a closer
relationship in the future where the spatial
fidelity of TEM is combined with the
analytical sensitivity and three
dimensionality of APT Instruments that
combine these techniques into a single
instrument are currently being pursued by
adding a STEM to a LEAP [22] and a LEAP
to a STEM [23]
This work demonstrates that a wealth
of high quality information may be obtained
Figure 3 APT atom map containing
As (large black spheres) B (small
dark grey spheres) and HfO (small
light grey spheres)
in figure 3 which is an atom map containing As (large black
spheres) B (small dark grey spheres) and HfO (small light
grey spheres) together with a 12at Ge isoconcentration
surface (arrowed) The Hf is detected entirely in HfO complex
molecule peaks shown in figure 4 which are detected in the
2+ charge state over the range of 95 to 99 Da Meas
peak at 98 Da the mass resolving power is gt1000 (FWHM)
which demonstrates the capability to achieve good spectral resolution even on
very complicated heterogeneous structures The Hf atoms are detected at the top of the image
3 in a region estimated to have a composition of approximately 80at (Hf+O) and 20at Si
that the apparent HfO molecules detected in the channel region are spectral noise due to the large
number of mass ranges required for the HfO isotopes) The maximum level of Ge (positioned along
the left and right edges of the image shown in figure 3 is ~25at The Ge isoconcentration surface
delineates the sourcedrain regions from the channel region
ch resembles the undercut shape of the channel region near the gate oxide shown in figure
This shape correlation may be used to create a composite image of the STEM and APT data
he APT image was scaled to match the STEM image Note the qualitative evidence
of carbon atoms clustering together in the lower center portion of the image The estimated
e APT data is
Although these data are not from
exactly the same volume of material
very few
[21]) the
exercise is still useful It provides us with
information on the accuracy of the APT data
s well as perhaps a glimpse
microscopy
Indeed TEM and APT are complementary
techniques that will likely develop a closer
relationship in the future where the spatial
fidelity of TEM is combined with the
lytical sensitivity and three
dimensionality of APT Instruments that
combine these techniques into a single
instrument are currently being pursued by
STEM to a LEAP [22] and a LEAP
This work demonstrates that a wealth
of high quality information may be obtained
APT atom map containing
dark grey spheres) and HfO (small
Figure 4 HfO complex molecule peaks detected in the 2+
charge state over the range of 95 to 99 Da The inset table shows
the comparison of expected isotopic abundances (using gt02at
only) to the detected abundances
Figure 5 Superimposed STEM image and APT atom
map containing B (small dark grey spheres) and
carbon (large light grey spheres) atoms only Note the
indications of carbon atoms clustering together (The
ATP data are 20nm in thickness into the page)
in figure 3 which is an atom map containing As (large black
s) and HfO (small light
12at Ge isoconcentration
The Hf is detected entirely in HfO complex
detected in the
Measured at the
the mass resolving power is gt1000 (FWHM)
the capability to achieve good spectral resolution even on
the top of the image in figure
f+O) and 20at Si (Note
that the apparent HfO molecules detected in the channel region are spectral noise due to the large
The maximum level of Ge (positioned along
The Ge isoconcentration surface
the sourcedrain regions from the channel region and has a
ch resembles the undercut shape of the channel region near the gate oxide shown in figure
ite image of the STEM and APT data
Note the qualitative evidence
in the lower center portion of the image The estimated
complex molecule peaks detected in the 2+
charge state over the range of 95 to 99 Da The inset table shows
the comparison of expected isotopic abundances (using gt02at
Superimposed STEM image and APT atom
ap containing B (small dark grey spheres) and
carbon (large light grey spheres) atoms only Note the
indications of carbon atoms clustering together (The
ATP data are 20nm in thickness into the page)
17th International Conference on Microscopy of Semiconducting Materials 2011 IOP PublishingJournal of Physics Conference Series 326 (2011) 012030 doi1010881742-65963261012030
3
from site-specific atom probe analysis of post-production microelectronic devices Adequate yields
(gt50) need to be realized and APT reconstruction methods improved going forward but certainly at
this time feasibility has been shown
Acknowledgements
The authors would like to thank our colleagues at Cameca Instruments Inc who assisted in
assembling the materials presented in this manuscript including B Geiser J Olson J Shepard T
Payne E Strennen E Oltman T Gribb D Rauls J Watson and S Gerstl (currently at
Eidgenoumlssische Technische Hochschule Zuumlrich) We would especially like to thank P Ronsheim
(IBM) for helpful discussions
References
[1] International Technology Roadmap for Semiconductors (httpwwwitrsnet) [2] Muumlller E W Panitz J A and McLane S B 1968 Review of Scientific Instruments 39 83-6 [3] Kellogg G L and Tsong T T 1980 Journal of Applied Physics 51 1184-94 [4] Kelly T F Larson D J Thompson K Alvis R L Bunton J H Olson J D and Gorman B P 2007
Annual Review of Materials Research 37 681-727 [5] Lauhon L J Adusumilli P Ronsheim P Flaitz P L and Lawrence D 2009 MRS Bulletin 34 738-
43 [6] Larson D J Prosa T J Lawrence D Geiser B P Jones C M and Kelly T F 2011 Handbook of
Instrumentation and Techniques for Semiconductor Nanostructure Characterization ed R Haight et al (London World Scientific PublishingImperial College Press)
[7] Mutas S Klein C and Gerstl S S A 2011 Ultramicroscopy in press [8] Larson D J Alvis R A Lawrence D F Prosa T J Ulfig R M Reinhard D A Clifton P H Gerstl
S S A Bunton J H Lenz D R Kelly T F and Stiller K 2008 Microscopy and Microanalysis 14 1254-5
[9] Chen Y M Ohkubo T Kodzuka M Morita K and Hono K 2009 Scripta Materialia 61 693ndash6 [10] Marquis E A Yahya N A Larson D J Miller M K and Todd R I 2010 Materials Today 13(10)
42-4 [11] Li F Ohkubo T Chen Y M Kodzuka M Ye F Ou D R Mori T and Hono K 2010 Scripta
Materialia 63 332-5 [12] Payne D J and Marquis E A 2011 Chemistry of Materials 23 1085-7 [13] Moore J S Jones K S Kennel H and Corcoran S 2008 Ultramicroscopy 108 536ndash9 [14] Inoue K Yano F Nishida A Takamizawa H Tsunomura T Nagai Y and Hasegawa M 2009
Ultramicroscopy 109 1479-84 [15] Kambham A K Mody J Gilbert M Koelling S and Vandervorst W 2010 Ultramicroscopy in
press [16] Miller M K Russell K F Thompson K Alvis R and Larson D J 2007 Microscopy and
Microanalysis 13 428-36 [17] Marquis E A Geiser B P Prosa T J and Larson D J 2011 Journal of Microscopy 241 255 [18] SVTC Technologies (httpwwwsvtccom) [19] Thompson K Lawrence D J Larson D J Olson J D Kelly T F and Gorman B 2007
Ultramicroscopy 107 131-9 [20] Larson D J Foord D T Petford-Long A K Liew H Blamire M G Cerezo A and Smith G D W
1999 Ultramicroscopy 79 287-93 [21] Arslan I Marquis E A Homer M Hekmaty M A and Bartelt N C 2008 Ultramicroscopy 108
1579-1585 [22] Gorman B P et al Microscopy and Microanalysis (2011) Supplement 2 in press [23] Kelly T F et al Microscopy and Microanalysis (2011) Supplement 2 in press
17th International Conference on Microscopy of Semiconducting Materials 2011 IOP PublishingJournal of Physics Conference Series 326 (2011) 012030 doi1010881742-65963261012030
4
Toward atom probe tomography of microelectronic devices
D J Larson1
D Lawrence1 W Lefebvre
2 D Olson
1 T J Prosa
1 D A Reinhard
1
R M Ulfig1 P H Clifton
1 J H Bunton
1 D Lenz
1 J D Olson
1 L Renaud
3
I Martin3 and T F Kelly
1
1 Cameca Instruments Inc 5500 Nobel Drive Madison WI 53711 USA
2 Universiteacute de Rouen Saint Etienne du Rouvray 76801 FRANCE
3 Cameca SAS 29 Quai des Greacutesillons Gennevilliers 92622 FRANCE
Summary Atom probe tomography and scanning transmission electron microscopy has been
used to analyze a commercial microelectronics device prepared by depackaging and focused
ion beam milling Chemical and morphological data are presented from the source drain and
channel regions and part of the gate oxide region of an Intel
i5-650 p-FET device
demonstrating feasibility in using these techniques to investigate commercial chips
1 Introduction
Shrinking feature sizes continue to provide one of the main challenges for physical metrology
methods From the 2009 ITRS [1] ldquoTo achieve desired device scaling metrology tools must be
capable of measurement of properties on atomic distancesrdquo Atom probe tomography nominally has
this capability and has been used in materials characterization and materials science for more than 40
years [2] Although using a laser to assist field evaporation of materials in the atom probe was
developed ~30 years ago [3] only recently has APT begun to see widespread usage in the areas of
semiconductors [4-7] and ceramics [8-12] Although APT has also recently been applied to transistor
and FINFET-type structures [13-15] these structures usually have been stopped at some point in the
semiconductor build process in order to accommodate the APT analysis As it is a complex multi-step
task to analyze a fully processed microelectronic device using APT we will take a step back and
evaluate the three steps that are necessary to result in a successful APT analysis of any structure
Number one is specimen preparation For APT the use of focused ion beam (FIB) instruments
has revolutionized specimen preparation (for a recent review see [16]) For the case of microelectronic
device analysis FIB preparation to isolate lt50nm heterogeneous devices in arbitrary XYZ
orientations is critical This occurs only after suitable deprocessing (may consist of depackaging dry
and wet electrochemistry for selective material etching etc) has been successfully achieved Since a
protective layer is needed if near-surface features are to be analyzed proper ldquocap matchingrdquo becomes
important The important requirements are relatively low temperature deposition (to avoid inducing
undesired diffusion) matching of evaporation field with target material (to minimize reconstruction
artifacts) and good adhesion
Number two is specimen yield during data collection Even though a specimen may be prepared
which contains the region of interest specimen yield must be adequate (for the individual userrsquos
purpose or specific laboratory environment) Clearly the details of specimen preparation are
intricately related to yield but data collection parameters may also have a significant effect on yield
and there is often a trade-off between yield and data quality (primarily mass spectral and background
noise quality)
Davidlarsonametekcom
17th International Conference on Microscopy of Semiconducting Materials 2011 IOP PublishingJournal of Physics Conference Series 326 (2011) 012030 doi1010881742-65963261012030
Published under licence by IOP Publishing Ltd 1
Number three is
APT data
reconstruction which
is not discussed
significantly in this
document Data may
be collected with an
adequate yield but if
the resulting data
reconstruction is either
not accurate or not
precise enough then
the analysis may fail to
meet requirements
The main problem
with APT data
reconstruction from
the perspective of devices is that field evaporation of heterogeneous struc
evaporated surfaces that are far from hemispherical
fundamental assumption employed in the majority
This work presents the status
commercial microelectronics device (
we acknowledge that satisfactory yield has not
reconstruction of transistors are non
2 Experimental
Specimen preparation was carried out in an FEI Novalab
with an Omniprobe AP200 in-situ
[110] orientation of the Si substrate with a JEM ARM 200F operating at 200kV equipped with a
Schottky field emission gun APT data collection was performed on a LEAP
Cameca Instruments Inc The atom probe was operated in a 200 kHz pulsed laser mode with an
energy of 100 pJ into an estimated spot size (four sigma) at the specimen of ~3
temperature was 50K and the ion detection rate was 020 (1 ion detected in every 500 laser pulses)
The specimen examined in this work is
(Intel i5-650) which was purchased at a
retail outlet Following depackagi
(approximately) metal-1 [18] focuse
ion-beam preparation [19] was used to
create specimens Figure 1 shows
sequence of images throughout the
process from the initial surface of the
wafer after depackaging (figure 1a)
through to the final state of focused
beam annular sharpening [20] of the tip
(figure 1h)
3 Results and Discussion
The device structure analyzed with APT
is shown in figure 2 which contains a bright
figure 2a a multilayered gate oxide structure is
angled regions between the SiGeB sourcedrain regions and the channel become much more obvious
The APT analysis obtained from a device in a region near the one
Figure 1
APT analysis a) initial su
coupon d)
beam annular sharpening
is that field evaporation of heterogeneous structures often leads to
are far from hemispherical [17] and a hemispherical end form
assumption employed in the majority of current reconstruction algorithms
This work presents the status of efforts to prepare analyze and reconstruct data from
commercial microelectronics device (32 nm node Intel
i5-650 nFET) This is a work in progress and
we acknowledge that satisfactory yield has not yet been obtained Also issues relating to APT data
of transistors are non-trivial and will not be discussed further herein
Specimen preparation was carried out in an FEI Novalab dual-beam focused ion beam instrument
micromanipulator STEM observations were performed along the
e Si substrate with a JEM ARM 200F operating at 200kV equipped with a
APT data collection was performed on a LEAP 4000XH
The atom probe was operated in a 200 kHz pulsed laser mode with an
of 100 pJ into an estimated spot size (four sigma) at the specimen of ~3 microm The specimen
temperature was 50K and the ion detection rate was 020 (1 ion detected in every 500 laser pulses)
The specimen examined in this work is a device from a commercial 32-nm technology
was purchased at a
retail outlet Following depackaging to
focused-
was used to
1 shows a
sequence of images throughout the
the initial surface of the
(figure 1a)
focused-ion-
of the tip
with APT
ntains a bright-field high-angle annular-dark-field pair of images
gate oxide structure is clearly visible while in figure 2b the
sourcedrain regions and the channel become much more obvious
The APT analysis obtained from a device in a region near the one shown in figure 2 is
Focused ion beam preparation of a microelectronic device for
APT analysis a) initial surface b) FIB-milled trenches c) the extracted
coupon d) pre-tip cut from the coupon and e-h) the stages of focused
beam annular sharpening
Figure 2 (a) Bright-field and (b) high-angle annular
field images of the Intel i5-650 device
tures often leads to
hemispherical end form is a
and reconstruct data from a
650 nFET) This is a work in progress and
issues relating to APT data
focused ion beam instrument
STEM observations were performed along the
e Si substrate with a JEM ARM 200F operating at 200kV equipped with a
4000XHR from
The atom probe was operated in a 200 kHz pulsed laser mode with an
m The specimen
temperature was 50K and the ion detection rate was 020 (1 ion detected in every 500 laser pulses)
nm technology chip
field pair of images In
visible while in figure 2b the undercut and
sourcedrain regions and the channel become much more obvious
in figure 2 is presented
Focused ion beam preparation of a microelectronic device for
milled trenches c) the extracted
stages of focused-ion-
angle annular-dark-
17th International Conference on Microscopy of Semiconducting Materials 2011 IOP PublishingJournal of Physics Conference Series 326 (2011) 012030 doi1010881742-65963261012030
2
and gt500 (FWTM) which demonstrates
very complicated heterogeneous structures The Hf
3 in a region estimated to have a composition of approximately 80at (H
that the apparent HfO molecules detected in the channel region are spectral noise due to the large
number of mass ranges required for the HfO isot
the left and right edges of the image shown in figure 3 is ~25at
shown in figure 3 (arrowed) delineate
shape which resembles the undercut shape of the channel region near the gate oxide shown in figure
2b
This shape correlation may be used to create a compos
figure 5 in which the APT image was scaled to match the STEM image
of carbon atoms clustering together
concentration of carbon in the APT data is
~010at
Although these data are not from
exactly the same volume of material
(something which has been done very few
times to date in the literature [21]
exercise is still useful It provides us with
information on the accuracy of the APT da
reconstruction as well as perhaps a
of the future of correlative microscopy
Indeed TEM and APT are complementary
techniques that will likely develop a closer
relationship in the future where the spatial
fidelity of TEM is combined with the
analytical sensitivity and three
dimensionality of APT Instruments that
combine these techniques into a single
instrument are currently being pursued by
adding a STEM to a LEAP [22] and a LEAP
to a STEM [23]
This work demonstrates that a wealth
of high quality information may be obtained
Figure 3 APT atom map containing
As (large black spheres) B (small
dark grey spheres) and HfO (small
light grey spheres)
in figure 3 which is an atom map containing As (large black
spheres) B (small dark grey spheres) and HfO (small light
grey spheres) together with a 12at Ge isoconcentration
surface (arrowed) The Hf is detected entirely in HfO complex
molecule peaks shown in figure 4 which are detected in the
2+ charge state over the range of 95 to 99 Da Meas
peak at 98 Da the mass resolving power is gt1000 (FWHM)
which demonstrates the capability to achieve good spectral resolution even on
very complicated heterogeneous structures The Hf atoms are detected at the top of the image
3 in a region estimated to have a composition of approximately 80at (Hf+O) and 20at Si
that the apparent HfO molecules detected in the channel region are spectral noise due to the large
number of mass ranges required for the HfO isotopes) The maximum level of Ge (positioned along
the left and right edges of the image shown in figure 3 is ~25at The Ge isoconcentration surface
delineates the sourcedrain regions from the channel region
ch resembles the undercut shape of the channel region near the gate oxide shown in figure
This shape correlation may be used to create a composite image of the STEM and APT data
he APT image was scaled to match the STEM image Note the qualitative evidence
of carbon atoms clustering together in the lower center portion of the image The estimated
e APT data is
Although these data are not from
exactly the same volume of material
very few
[21]) the
exercise is still useful It provides us with
information on the accuracy of the APT data
s well as perhaps a glimpse
microscopy
Indeed TEM and APT are complementary
techniques that will likely develop a closer
relationship in the future where the spatial
fidelity of TEM is combined with the
lytical sensitivity and three
dimensionality of APT Instruments that
combine these techniques into a single
instrument are currently being pursued by
STEM to a LEAP [22] and a LEAP
This work demonstrates that a wealth
of high quality information may be obtained
APT atom map containing
dark grey spheres) and HfO (small
Figure 4 HfO complex molecule peaks detected in the 2+
charge state over the range of 95 to 99 Da The inset table shows
the comparison of expected isotopic abundances (using gt02at
only) to the detected abundances
Figure 5 Superimposed STEM image and APT atom
map containing B (small dark grey spheres) and
carbon (large light grey spheres) atoms only Note the
indications of carbon atoms clustering together (The
ATP data are 20nm in thickness into the page)
in figure 3 which is an atom map containing As (large black
s) and HfO (small light
12at Ge isoconcentration
The Hf is detected entirely in HfO complex
detected in the
Measured at the
the mass resolving power is gt1000 (FWHM)
the capability to achieve good spectral resolution even on
the top of the image in figure
f+O) and 20at Si (Note
that the apparent HfO molecules detected in the channel region are spectral noise due to the large
The maximum level of Ge (positioned along
The Ge isoconcentration surface
the sourcedrain regions from the channel region and has a
ch resembles the undercut shape of the channel region near the gate oxide shown in figure
ite image of the STEM and APT data
Note the qualitative evidence
in the lower center portion of the image The estimated
complex molecule peaks detected in the 2+
charge state over the range of 95 to 99 Da The inset table shows
the comparison of expected isotopic abundances (using gt02at
Superimposed STEM image and APT atom
ap containing B (small dark grey spheres) and
carbon (large light grey spheres) atoms only Note the
indications of carbon atoms clustering together (The
ATP data are 20nm in thickness into the page)
17th International Conference on Microscopy of Semiconducting Materials 2011 IOP PublishingJournal of Physics Conference Series 326 (2011) 012030 doi1010881742-65963261012030
3
from site-specific atom probe analysis of post-production microelectronic devices Adequate yields
(gt50) need to be realized and APT reconstruction methods improved going forward but certainly at
this time feasibility has been shown
Acknowledgements
The authors would like to thank our colleagues at Cameca Instruments Inc who assisted in
assembling the materials presented in this manuscript including B Geiser J Olson J Shepard T
Payne E Strennen E Oltman T Gribb D Rauls J Watson and S Gerstl (currently at
Eidgenoumlssische Technische Hochschule Zuumlrich) We would especially like to thank P Ronsheim
(IBM) for helpful discussions
References
[1] International Technology Roadmap for Semiconductors (httpwwwitrsnet) [2] Muumlller E W Panitz J A and McLane S B 1968 Review of Scientific Instruments 39 83-6 [3] Kellogg G L and Tsong T T 1980 Journal of Applied Physics 51 1184-94 [4] Kelly T F Larson D J Thompson K Alvis R L Bunton J H Olson J D and Gorman B P 2007
Annual Review of Materials Research 37 681-727 [5] Lauhon L J Adusumilli P Ronsheim P Flaitz P L and Lawrence D 2009 MRS Bulletin 34 738-
43 [6] Larson D J Prosa T J Lawrence D Geiser B P Jones C M and Kelly T F 2011 Handbook of
Instrumentation and Techniques for Semiconductor Nanostructure Characterization ed R Haight et al (London World Scientific PublishingImperial College Press)
[7] Mutas S Klein C and Gerstl S S A 2011 Ultramicroscopy in press [8] Larson D J Alvis R A Lawrence D F Prosa T J Ulfig R M Reinhard D A Clifton P H Gerstl
S S A Bunton J H Lenz D R Kelly T F and Stiller K 2008 Microscopy and Microanalysis 14 1254-5
[9] Chen Y M Ohkubo T Kodzuka M Morita K and Hono K 2009 Scripta Materialia 61 693ndash6 [10] Marquis E A Yahya N A Larson D J Miller M K and Todd R I 2010 Materials Today 13(10)
42-4 [11] Li F Ohkubo T Chen Y M Kodzuka M Ye F Ou D R Mori T and Hono K 2010 Scripta
Materialia 63 332-5 [12] Payne D J and Marquis E A 2011 Chemistry of Materials 23 1085-7 [13] Moore J S Jones K S Kennel H and Corcoran S 2008 Ultramicroscopy 108 536ndash9 [14] Inoue K Yano F Nishida A Takamizawa H Tsunomura T Nagai Y and Hasegawa M 2009
Ultramicroscopy 109 1479-84 [15] Kambham A K Mody J Gilbert M Koelling S and Vandervorst W 2010 Ultramicroscopy in
press [16] Miller M K Russell K F Thompson K Alvis R and Larson D J 2007 Microscopy and
Microanalysis 13 428-36 [17] Marquis E A Geiser B P Prosa T J and Larson D J 2011 Journal of Microscopy 241 255 [18] SVTC Technologies (httpwwwsvtccom) [19] Thompson K Lawrence D J Larson D J Olson J D Kelly T F and Gorman B 2007
Ultramicroscopy 107 131-9 [20] Larson D J Foord D T Petford-Long A K Liew H Blamire M G Cerezo A and Smith G D W
1999 Ultramicroscopy 79 287-93 [21] Arslan I Marquis E A Homer M Hekmaty M A and Bartelt N C 2008 Ultramicroscopy 108
1579-1585 [22] Gorman B P et al Microscopy and Microanalysis (2011) Supplement 2 in press [23] Kelly T F et al Microscopy and Microanalysis (2011) Supplement 2 in press
17th International Conference on Microscopy of Semiconducting Materials 2011 IOP PublishingJournal of Physics Conference Series 326 (2011) 012030 doi1010881742-65963261012030
4
Number three is
APT data
reconstruction which
is not discussed
significantly in this
document Data may
be collected with an
adequate yield but if
the resulting data
reconstruction is either
not accurate or not
precise enough then
the analysis may fail to
meet requirements
The main problem
with APT data
reconstruction from
the perspective of devices is that field evaporation of heterogeneous struc
evaporated surfaces that are far from hemispherical
fundamental assumption employed in the majority
This work presents the status
commercial microelectronics device (
we acknowledge that satisfactory yield has not
reconstruction of transistors are non
2 Experimental
Specimen preparation was carried out in an FEI Novalab
with an Omniprobe AP200 in-situ
[110] orientation of the Si substrate with a JEM ARM 200F operating at 200kV equipped with a
Schottky field emission gun APT data collection was performed on a LEAP
Cameca Instruments Inc The atom probe was operated in a 200 kHz pulsed laser mode with an
energy of 100 pJ into an estimated spot size (four sigma) at the specimen of ~3
temperature was 50K and the ion detection rate was 020 (1 ion detected in every 500 laser pulses)
The specimen examined in this work is
(Intel i5-650) which was purchased at a
retail outlet Following depackagi
(approximately) metal-1 [18] focuse
ion-beam preparation [19] was used to
create specimens Figure 1 shows
sequence of images throughout the
process from the initial surface of the
wafer after depackaging (figure 1a)
through to the final state of focused
beam annular sharpening [20] of the tip
(figure 1h)
3 Results and Discussion
The device structure analyzed with APT
is shown in figure 2 which contains a bright
figure 2a a multilayered gate oxide structure is
angled regions between the SiGeB sourcedrain regions and the channel become much more obvious
The APT analysis obtained from a device in a region near the one
Figure 1
APT analysis a) initial su
coupon d)
beam annular sharpening
is that field evaporation of heterogeneous structures often leads to
are far from hemispherical [17] and a hemispherical end form
assumption employed in the majority of current reconstruction algorithms
This work presents the status of efforts to prepare analyze and reconstruct data from
commercial microelectronics device (32 nm node Intel
i5-650 nFET) This is a work in progress and
we acknowledge that satisfactory yield has not yet been obtained Also issues relating to APT data
of transistors are non-trivial and will not be discussed further herein
Specimen preparation was carried out in an FEI Novalab dual-beam focused ion beam instrument
micromanipulator STEM observations were performed along the
e Si substrate with a JEM ARM 200F operating at 200kV equipped with a
APT data collection was performed on a LEAP 4000XH
The atom probe was operated in a 200 kHz pulsed laser mode with an
of 100 pJ into an estimated spot size (four sigma) at the specimen of ~3 microm The specimen
temperature was 50K and the ion detection rate was 020 (1 ion detected in every 500 laser pulses)
The specimen examined in this work is a device from a commercial 32-nm technology
was purchased at a
retail outlet Following depackaging to
focused-
was used to
1 shows a
sequence of images throughout the
the initial surface of the
(figure 1a)
focused-ion-
of the tip
with APT
ntains a bright-field high-angle annular-dark-field pair of images
gate oxide structure is clearly visible while in figure 2b the
sourcedrain regions and the channel become much more obvious
The APT analysis obtained from a device in a region near the one shown in figure 2 is
Focused ion beam preparation of a microelectronic device for
APT analysis a) initial surface b) FIB-milled trenches c) the extracted
coupon d) pre-tip cut from the coupon and e-h) the stages of focused
beam annular sharpening
Figure 2 (a) Bright-field and (b) high-angle annular
field images of the Intel i5-650 device
tures often leads to
hemispherical end form is a
and reconstruct data from a
650 nFET) This is a work in progress and
issues relating to APT data
focused ion beam instrument
STEM observations were performed along the
e Si substrate with a JEM ARM 200F operating at 200kV equipped with a
4000XHR from
The atom probe was operated in a 200 kHz pulsed laser mode with an
m The specimen
temperature was 50K and the ion detection rate was 020 (1 ion detected in every 500 laser pulses)
nm technology chip
field pair of images In
visible while in figure 2b the undercut and
sourcedrain regions and the channel become much more obvious
in figure 2 is presented
Focused ion beam preparation of a microelectronic device for
milled trenches c) the extracted
stages of focused-ion-
angle annular-dark-
17th International Conference on Microscopy of Semiconducting Materials 2011 IOP PublishingJournal of Physics Conference Series 326 (2011) 012030 doi1010881742-65963261012030
2
and gt500 (FWTM) which demonstrates
very complicated heterogeneous structures The Hf
3 in a region estimated to have a composition of approximately 80at (H
that the apparent HfO molecules detected in the channel region are spectral noise due to the large
number of mass ranges required for the HfO isot
the left and right edges of the image shown in figure 3 is ~25at
shown in figure 3 (arrowed) delineate
shape which resembles the undercut shape of the channel region near the gate oxide shown in figure
2b
This shape correlation may be used to create a compos
figure 5 in which the APT image was scaled to match the STEM image
of carbon atoms clustering together
concentration of carbon in the APT data is
~010at
Although these data are not from
exactly the same volume of material
(something which has been done very few
times to date in the literature [21]
exercise is still useful It provides us with
information on the accuracy of the APT da
reconstruction as well as perhaps a
of the future of correlative microscopy
Indeed TEM and APT are complementary
techniques that will likely develop a closer
relationship in the future where the spatial
fidelity of TEM is combined with the
analytical sensitivity and three
dimensionality of APT Instruments that
combine these techniques into a single
instrument are currently being pursued by
adding a STEM to a LEAP [22] and a LEAP
to a STEM [23]
This work demonstrates that a wealth
of high quality information may be obtained
Figure 3 APT atom map containing
As (large black spheres) B (small
dark grey spheres) and HfO (small
light grey spheres)
in figure 3 which is an atom map containing As (large black
spheres) B (small dark grey spheres) and HfO (small light
grey spheres) together with a 12at Ge isoconcentration
surface (arrowed) The Hf is detected entirely in HfO complex
molecule peaks shown in figure 4 which are detected in the
2+ charge state over the range of 95 to 99 Da Meas
peak at 98 Da the mass resolving power is gt1000 (FWHM)
which demonstrates the capability to achieve good spectral resolution even on
very complicated heterogeneous structures The Hf atoms are detected at the top of the image
3 in a region estimated to have a composition of approximately 80at (Hf+O) and 20at Si
that the apparent HfO molecules detected in the channel region are spectral noise due to the large
number of mass ranges required for the HfO isotopes) The maximum level of Ge (positioned along
the left and right edges of the image shown in figure 3 is ~25at The Ge isoconcentration surface
delineates the sourcedrain regions from the channel region
ch resembles the undercut shape of the channel region near the gate oxide shown in figure
This shape correlation may be used to create a composite image of the STEM and APT data
he APT image was scaled to match the STEM image Note the qualitative evidence
of carbon atoms clustering together in the lower center portion of the image The estimated
e APT data is
Although these data are not from
exactly the same volume of material
very few
[21]) the
exercise is still useful It provides us with
information on the accuracy of the APT data
s well as perhaps a glimpse
microscopy
Indeed TEM and APT are complementary
techniques that will likely develop a closer
relationship in the future where the spatial
fidelity of TEM is combined with the
lytical sensitivity and three
dimensionality of APT Instruments that
combine these techniques into a single
instrument are currently being pursued by
STEM to a LEAP [22] and a LEAP
This work demonstrates that a wealth
of high quality information may be obtained
APT atom map containing
dark grey spheres) and HfO (small
Figure 4 HfO complex molecule peaks detected in the 2+
charge state over the range of 95 to 99 Da The inset table shows
the comparison of expected isotopic abundances (using gt02at
only) to the detected abundances
Figure 5 Superimposed STEM image and APT atom
map containing B (small dark grey spheres) and
carbon (large light grey spheres) atoms only Note the
indications of carbon atoms clustering together (The
ATP data are 20nm in thickness into the page)
in figure 3 which is an atom map containing As (large black
s) and HfO (small light
12at Ge isoconcentration
The Hf is detected entirely in HfO complex
detected in the
Measured at the
the mass resolving power is gt1000 (FWHM)
the capability to achieve good spectral resolution even on
the top of the image in figure
f+O) and 20at Si (Note
that the apparent HfO molecules detected in the channel region are spectral noise due to the large
The maximum level of Ge (positioned along
The Ge isoconcentration surface
the sourcedrain regions from the channel region and has a
ch resembles the undercut shape of the channel region near the gate oxide shown in figure
ite image of the STEM and APT data
Note the qualitative evidence
in the lower center portion of the image The estimated
complex molecule peaks detected in the 2+
charge state over the range of 95 to 99 Da The inset table shows
the comparison of expected isotopic abundances (using gt02at
Superimposed STEM image and APT atom
ap containing B (small dark grey spheres) and
carbon (large light grey spheres) atoms only Note the
indications of carbon atoms clustering together (The
ATP data are 20nm in thickness into the page)
17th International Conference on Microscopy of Semiconducting Materials 2011 IOP PublishingJournal of Physics Conference Series 326 (2011) 012030 doi1010881742-65963261012030
3
from site-specific atom probe analysis of post-production microelectronic devices Adequate yields
(gt50) need to be realized and APT reconstruction methods improved going forward but certainly at
this time feasibility has been shown
Acknowledgements
The authors would like to thank our colleagues at Cameca Instruments Inc who assisted in
assembling the materials presented in this manuscript including B Geiser J Olson J Shepard T
Payne E Strennen E Oltman T Gribb D Rauls J Watson and S Gerstl (currently at
Eidgenoumlssische Technische Hochschule Zuumlrich) We would especially like to thank P Ronsheim
(IBM) for helpful discussions
References
[1] International Technology Roadmap for Semiconductors (httpwwwitrsnet) [2] Muumlller E W Panitz J A and McLane S B 1968 Review of Scientific Instruments 39 83-6 [3] Kellogg G L and Tsong T T 1980 Journal of Applied Physics 51 1184-94 [4] Kelly T F Larson D J Thompson K Alvis R L Bunton J H Olson J D and Gorman B P 2007
Annual Review of Materials Research 37 681-727 [5] Lauhon L J Adusumilli P Ronsheim P Flaitz P L and Lawrence D 2009 MRS Bulletin 34 738-
43 [6] Larson D J Prosa T J Lawrence D Geiser B P Jones C M and Kelly T F 2011 Handbook of
Instrumentation and Techniques for Semiconductor Nanostructure Characterization ed R Haight et al (London World Scientific PublishingImperial College Press)
[7] Mutas S Klein C and Gerstl S S A 2011 Ultramicroscopy in press [8] Larson D J Alvis R A Lawrence D F Prosa T J Ulfig R M Reinhard D A Clifton P H Gerstl
S S A Bunton J H Lenz D R Kelly T F and Stiller K 2008 Microscopy and Microanalysis 14 1254-5
[9] Chen Y M Ohkubo T Kodzuka M Morita K and Hono K 2009 Scripta Materialia 61 693ndash6 [10] Marquis E A Yahya N A Larson D J Miller M K and Todd R I 2010 Materials Today 13(10)
42-4 [11] Li F Ohkubo T Chen Y M Kodzuka M Ye F Ou D R Mori T and Hono K 2010 Scripta
Materialia 63 332-5 [12] Payne D J and Marquis E A 2011 Chemistry of Materials 23 1085-7 [13] Moore J S Jones K S Kennel H and Corcoran S 2008 Ultramicroscopy 108 536ndash9 [14] Inoue K Yano F Nishida A Takamizawa H Tsunomura T Nagai Y and Hasegawa M 2009
Ultramicroscopy 109 1479-84 [15] Kambham A K Mody J Gilbert M Koelling S and Vandervorst W 2010 Ultramicroscopy in
press [16] Miller M K Russell K F Thompson K Alvis R and Larson D J 2007 Microscopy and
Microanalysis 13 428-36 [17] Marquis E A Geiser B P Prosa T J and Larson D J 2011 Journal of Microscopy 241 255 [18] SVTC Technologies (httpwwwsvtccom) [19] Thompson K Lawrence D J Larson D J Olson J D Kelly T F and Gorman B 2007
Ultramicroscopy 107 131-9 [20] Larson D J Foord D T Petford-Long A K Liew H Blamire M G Cerezo A and Smith G D W
1999 Ultramicroscopy 79 287-93 [21] Arslan I Marquis E A Homer M Hekmaty M A and Bartelt N C 2008 Ultramicroscopy 108
1579-1585 [22] Gorman B P et al Microscopy and Microanalysis (2011) Supplement 2 in press [23] Kelly T F et al Microscopy and Microanalysis (2011) Supplement 2 in press
17th International Conference on Microscopy of Semiconducting Materials 2011 IOP PublishingJournal of Physics Conference Series 326 (2011) 012030 doi1010881742-65963261012030
4
and gt500 (FWTM) which demonstrates
very complicated heterogeneous structures The Hf
3 in a region estimated to have a composition of approximately 80at (H
that the apparent HfO molecules detected in the channel region are spectral noise due to the large
number of mass ranges required for the HfO isot
the left and right edges of the image shown in figure 3 is ~25at
shown in figure 3 (arrowed) delineate
shape which resembles the undercut shape of the channel region near the gate oxide shown in figure
2b
This shape correlation may be used to create a compos
figure 5 in which the APT image was scaled to match the STEM image
of carbon atoms clustering together
concentration of carbon in the APT data is
~010at
Although these data are not from
exactly the same volume of material
(something which has been done very few
times to date in the literature [21]
exercise is still useful It provides us with
information on the accuracy of the APT da
reconstruction as well as perhaps a
of the future of correlative microscopy
Indeed TEM and APT are complementary
techniques that will likely develop a closer
relationship in the future where the spatial
fidelity of TEM is combined with the
analytical sensitivity and three
dimensionality of APT Instruments that
combine these techniques into a single
instrument are currently being pursued by
adding a STEM to a LEAP [22] and a LEAP
to a STEM [23]
This work demonstrates that a wealth
of high quality information may be obtained
Figure 3 APT atom map containing
As (large black spheres) B (small
dark grey spheres) and HfO (small
light grey spheres)
in figure 3 which is an atom map containing As (large black
spheres) B (small dark grey spheres) and HfO (small light
grey spheres) together with a 12at Ge isoconcentration
surface (arrowed) The Hf is detected entirely in HfO complex
molecule peaks shown in figure 4 which are detected in the
2+ charge state over the range of 95 to 99 Da Meas
peak at 98 Da the mass resolving power is gt1000 (FWHM)
which demonstrates the capability to achieve good spectral resolution even on
very complicated heterogeneous structures The Hf atoms are detected at the top of the image
3 in a region estimated to have a composition of approximately 80at (Hf+O) and 20at Si
that the apparent HfO molecules detected in the channel region are spectral noise due to the large
number of mass ranges required for the HfO isotopes) The maximum level of Ge (positioned along
the left and right edges of the image shown in figure 3 is ~25at The Ge isoconcentration surface
delineates the sourcedrain regions from the channel region
ch resembles the undercut shape of the channel region near the gate oxide shown in figure
This shape correlation may be used to create a composite image of the STEM and APT data
he APT image was scaled to match the STEM image Note the qualitative evidence
of carbon atoms clustering together in the lower center portion of the image The estimated
e APT data is
Although these data are not from
exactly the same volume of material
very few
[21]) the
exercise is still useful It provides us with
information on the accuracy of the APT data
s well as perhaps a glimpse
microscopy
Indeed TEM and APT are complementary
techniques that will likely develop a closer
relationship in the future where the spatial
fidelity of TEM is combined with the
lytical sensitivity and three
dimensionality of APT Instruments that
combine these techniques into a single
instrument are currently being pursued by
STEM to a LEAP [22] and a LEAP
This work demonstrates that a wealth
of high quality information may be obtained
APT atom map containing
dark grey spheres) and HfO (small
Figure 4 HfO complex molecule peaks detected in the 2+
charge state over the range of 95 to 99 Da The inset table shows
the comparison of expected isotopic abundances (using gt02at
only) to the detected abundances
Figure 5 Superimposed STEM image and APT atom
map containing B (small dark grey spheres) and
carbon (large light grey spheres) atoms only Note the
indications of carbon atoms clustering together (The
ATP data are 20nm in thickness into the page)
in figure 3 which is an atom map containing As (large black
s) and HfO (small light
12at Ge isoconcentration
The Hf is detected entirely in HfO complex
detected in the
Measured at the
the mass resolving power is gt1000 (FWHM)
the capability to achieve good spectral resolution even on
the top of the image in figure
f+O) and 20at Si (Note
that the apparent HfO molecules detected in the channel region are spectral noise due to the large
The maximum level of Ge (positioned along
The Ge isoconcentration surface
the sourcedrain regions from the channel region and has a
ch resembles the undercut shape of the channel region near the gate oxide shown in figure
ite image of the STEM and APT data
Note the qualitative evidence
in the lower center portion of the image The estimated
complex molecule peaks detected in the 2+
charge state over the range of 95 to 99 Da The inset table shows
the comparison of expected isotopic abundances (using gt02at
Superimposed STEM image and APT atom
ap containing B (small dark grey spheres) and
carbon (large light grey spheres) atoms only Note the
indications of carbon atoms clustering together (The
ATP data are 20nm in thickness into the page)
17th International Conference on Microscopy of Semiconducting Materials 2011 IOP PublishingJournal of Physics Conference Series 326 (2011) 012030 doi1010881742-65963261012030
3
from site-specific atom probe analysis of post-production microelectronic devices Adequate yields
(gt50) need to be realized and APT reconstruction methods improved going forward but certainly at
this time feasibility has been shown
Acknowledgements
The authors would like to thank our colleagues at Cameca Instruments Inc who assisted in
assembling the materials presented in this manuscript including B Geiser J Olson J Shepard T
Payne E Strennen E Oltman T Gribb D Rauls J Watson and S Gerstl (currently at
Eidgenoumlssische Technische Hochschule Zuumlrich) We would especially like to thank P Ronsheim
(IBM) for helpful discussions
References
[1] International Technology Roadmap for Semiconductors (httpwwwitrsnet) [2] Muumlller E W Panitz J A and McLane S B 1968 Review of Scientific Instruments 39 83-6 [3] Kellogg G L and Tsong T T 1980 Journal of Applied Physics 51 1184-94 [4] Kelly T F Larson D J Thompson K Alvis R L Bunton J H Olson J D and Gorman B P 2007
Annual Review of Materials Research 37 681-727 [5] Lauhon L J Adusumilli P Ronsheim P Flaitz P L and Lawrence D 2009 MRS Bulletin 34 738-
43 [6] Larson D J Prosa T J Lawrence D Geiser B P Jones C M and Kelly T F 2011 Handbook of
Instrumentation and Techniques for Semiconductor Nanostructure Characterization ed R Haight et al (London World Scientific PublishingImperial College Press)
[7] Mutas S Klein C and Gerstl S S A 2011 Ultramicroscopy in press [8] Larson D J Alvis R A Lawrence D F Prosa T J Ulfig R M Reinhard D A Clifton P H Gerstl
S S A Bunton J H Lenz D R Kelly T F and Stiller K 2008 Microscopy and Microanalysis 14 1254-5
[9] Chen Y M Ohkubo T Kodzuka M Morita K and Hono K 2009 Scripta Materialia 61 693ndash6 [10] Marquis E A Yahya N A Larson D J Miller M K and Todd R I 2010 Materials Today 13(10)
42-4 [11] Li F Ohkubo T Chen Y M Kodzuka M Ye F Ou D R Mori T and Hono K 2010 Scripta
Materialia 63 332-5 [12] Payne D J and Marquis E A 2011 Chemistry of Materials 23 1085-7 [13] Moore J S Jones K S Kennel H and Corcoran S 2008 Ultramicroscopy 108 536ndash9 [14] Inoue K Yano F Nishida A Takamizawa H Tsunomura T Nagai Y and Hasegawa M 2009
Ultramicroscopy 109 1479-84 [15] Kambham A K Mody J Gilbert M Koelling S and Vandervorst W 2010 Ultramicroscopy in
press [16] Miller M K Russell K F Thompson K Alvis R and Larson D J 2007 Microscopy and
Microanalysis 13 428-36 [17] Marquis E A Geiser B P Prosa T J and Larson D J 2011 Journal of Microscopy 241 255 [18] SVTC Technologies (httpwwwsvtccom) [19] Thompson K Lawrence D J Larson D J Olson J D Kelly T F and Gorman B 2007
Ultramicroscopy 107 131-9 [20] Larson D J Foord D T Petford-Long A K Liew H Blamire M G Cerezo A and Smith G D W
1999 Ultramicroscopy 79 287-93 [21] Arslan I Marquis E A Homer M Hekmaty M A and Bartelt N C 2008 Ultramicroscopy 108
1579-1585 [22] Gorman B P et al Microscopy and Microanalysis (2011) Supplement 2 in press [23] Kelly T F et al Microscopy and Microanalysis (2011) Supplement 2 in press
17th International Conference on Microscopy of Semiconducting Materials 2011 IOP PublishingJournal of Physics Conference Series 326 (2011) 012030 doi1010881742-65963261012030
4
from site-specific atom probe analysis of post-production microelectronic devices Adequate yields
(gt50) need to be realized and APT reconstruction methods improved going forward but certainly at
this time feasibility has been shown
Acknowledgements
The authors would like to thank our colleagues at Cameca Instruments Inc who assisted in
assembling the materials presented in this manuscript including B Geiser J Olson J Shepard T
Payne E Strennen E Oltman T Gribb D Rauls J Watson and S Gerstl (currently at
Eidgenoumlssische Technische Hochschule Zuumlrich) We would especially like to thank P Ronsheim
(IBM) for helpful discussions
References
[1] International Technology Roadmap for Semiconductors (httpwwwitrsnet) [2] Muumlller E W Panitz J A and McLane S B 1968 Review of Scientific Instruments 39 83-6 [3] Kellogg G L and Tsong T T 1980 Journal of Applied Physics 51 1184-94 [4] Kelly T F Larson D J Thompson K Alvis R L Bunton J H Olson J D and Gorman B P 2007
Annual Review of Materials Research 37 681-727 [5] Lauhon L J Adusumilli P Ronsheim P Flaitz P L and Lawrence D 2009 MRS Bulletin 34 738-
43 [6] Larson D J Prosa T J Lawrence D Geiser B P Jones C M and Kelly T F 2011 Handbook of
Instrumentation and Techniques for Semiconductor Nanostructure Characterization ed R Haight et al (London World Scientific PublishingImperial College Press)
[7] Mutas S Klein C and Gerstl S S A 2011 Ultramicroscopy in press [8] Larson D J Alvis R A Lawrence D F Prosa T J Ulfig R M Reinhard D A Clifton P H Gerstl
S S A Bunton J H Lenz D R Kelly T F and Stiller K 2008 Microscopy and Microanalysis 14 1254-5
[9] Chen Y M Ohkubo T Kodzuka M Morita K and Hono K 2009 Scripta Materialia 61 693ndash6 [10] Marquis E A Yahya N A Larson D J Miller M K and Todd R I 2010 Materials Today 13(10)
42-4 [11] Li F Ohkubo T Chen Y M Kodzuka M Ye F Ou D R Mori T and Hono K 2010 Scripta
Materialia 63 332-5 [12] Payne D J and Marquis E A 2011 Chemistry of Materials 23 1085-7 [13] Moore J S Jones K S Kennel H and Corcoran S 2008 Ultramicroscopy 108 536ndash9 [14] Inoue K Yano F Nishida A Takamizawa H Tsunomura T Nagai Y and Hasegawa M 2009
Ultramicroscopy 109 1479-84 [15] Kambham A K Mody J Gilbert M Koelling S and Vandervorst W 2010 Ultramicroscopy in
press [16] Miller M K Russell K F Thompson K Alvis R and Larson D J 2007 Microscopy and
Microanalysis 13 428-36 [17] Marquis E A Geiser B P Prosa T J and Larson D J 2011 Journal of Microscopy 241 255 [18] SVTC Technologies (httpwwwsvtccom) [19] Thompson K Lawrence D J Larson D J Olson J D Kelly T F and Gorman B 2007
Ultramicroscopy 107 131-9 [20] Larson D J Foord D T Petford-Long A K Liew H Blamire M G Cerezo A and Smith G D W
1999 Ultramicroscopy 79 287-93 [21] Arslan I Marquis E A Homer M Hekmaty M A and Bartelt N C 2008 Ultramicroscopy 108
1579-1585 [22] Gorman B P et al Microscopy and Microanalysis (2011) Supplement 2 in press [23] Kelly T F et al Microscopy and Microanalysis (2011) Supplement 2 in press
17th International Conference on Microscopy of Semiconducting Materials 2011 IOP PublishingJournal of Physics Conference Series 326 (2011) 012030 doi1010881742-65963261012030
4